Abstract

We recently proposed a new approach for the segmentation of speckled images based on active contours (snakes) [e.g., Opt.  Commun.   137, 382 (1997)]. We propose an extension of this approach to multichannel data. Two solutions are compared based on hypotheses on the possible mean intensity variation between the channels. Each solution is optimal for a certain class of input images, but one solution shows better or equivalent performance for both input image classes. This result opens new perspectives for the segmentation of multichannel images with the snake-based approach.

© 1999 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. J. W. Goodman, in Laser Speckle and Related PhenomenaJ. C. Dainty, ed., Vol.??9 of Topics in Applied Physics (Springer-Verlag, Berlin, 1975), pp. 9–75.
  2. J. Bruniquel and A. Lopes, Proc. SPIE 2315, 342 (1994).
    [CrossRef]
  3. A. Cumani, CVGIP Graph. Models Image Process. 53, 40 (1991).
    [CrossRef]
  4. P. E. Trahanias and A. N. Venetsanopoulos, IEEE Trans. Image Process. 2, 528 (1993).
    [CrossRef]
  5. O. Germain and Ph. Réfrégier, Opt. Lett. 22, 1845 (1996).
    [CrossRef]
  6. C. Chesnaud, V. Pagé, and Ph. Réfrégier, Opt. Lett. 7, 488 (1998).
    [CrossRef]
  7. Ph. Réfrégier, O. Germain, and T. Gaidon, Opt. Commun. 137, 382 (1997).
    [CrossRef]

1998

C. Chesnaud, V. Pagé, and Ph. Réfrégier, Opt. Lett. 7, 488 (1998).
[CrossRef]

1997

Ph. Réfrégier, O. Germain, and T. Gaidon, Opt. Commun. 137, 382 (1997).
[CrossRef]

1996

O. Germain and Ph. Réfrégier, Opt. Lett. 22, 1845 (1996).
[CrossRef]

1994

J. Bruniquel and A. Lopes, Proc. SPIE 2315, 342 (1994).
[CrossRef]

1993

P. E. Trahanias and A. N. Venetsanopoulos, IEEE Trans. Image Process. 2, 528 (1993).
[CrossRef]

1991

A. Cumani, CVGIP Graph. Models Image Process. 53, 40 (1991).
[CrossRef]

Bruniquel, J.

J. Bruniquel and A. Lopes, Proc. SPIE 2315, 342 (1994).
[CrossRef]

Chesnaud, C.

C. Chesnaud, V. Pagé, and Ph. Réfrégier, Opt. Lett. 7, 488 (1998).
[CrossRef]

Cumani, A.

A. Cumani, CVGIP Graph. Models Image Process. 53, 40 (1991).
[CrossRef]

Gaidon, T.

Ph. Réfrégier, O. Germain, and T. Gaidon, Opt. Commun. 137, 382 (1997).
[CrossRef]

Germain, O.

Ph. Réfrégier, O. Germain, and T. Gaidon, Opt. Commun. 137, 382 (1997).
[CrossRef]

O. Germain and Ph. Réfrégier, Opt. Lett. 22, 1845 (1996).
[CrossRef]

Goodman, J. W.

J. W. Goodman, in Laser Speckle and Related PhenomenaJ. C. Dainty, ed., Vol.??9 of Topics in Applied Physics (Springer-Verlag, Berlin, 1975), pp. 9–75.

Lopes, A.

J. Bruniquel and A. Lopes, Proc. SPIE 2315, 342 (1994).
[CrossRef]

Pagé, V.

C. Chesnaud, V. Pagé, and Ph. Réfrégier, Opt. Lett. 7, 488 (1998).
[CrossRef]

Réfrégier, Ph.

C. Chesnaud, V. Pagé, and Ph. Réfrégier, Opt. Lett. 7, 488 (1998).
[CrossRef]

Ph. Réfrégier, O. Germain, and T. Gaidon, Opt. Commun. 137, 382 (1997).
[CrossRef]

O. Germain and Ph. Réfrégier, Opt. Lett. 22, 1845 (1996).
[CrossRef]

Trahanias, P. E.

P. E. Trahanias and A. N. Venetsanopoulos, IEEE Trans. Image Process. 2, 528 (1993).
[CrossRef]

Venetsanopoulos, A. N.

P. E. Trahanias and A. N. Venetsanopoulos, IEEE Trans. Image Process. 2, 528 (1993).
[CrossRef]

CVGIP Graph. Models Image Process.

A. Cumani, CVGIP Graph. Models Image Process. 53, 40 (1991).
[CrossRef]

IEEE Trans. Image Process.

P. E. Trahanias and A. N. Venetsanopoulos, IEEE Trans. Image Process. 2, 528 (1993).
[CrossRef]

Opt. Commun.

Ph. Réfrégier, O. Germain, and T. Gaidon, Opt. Commun. 137, 382 (1997).
[CrossRef]

Opt. Lett.

O. Germain and Ph. Réfrégier, Opt. Lett. 22, 1845 (1996).
[CrossRef]

C. Chesnaud, V. Pagé, and Ph. Réfrégier, Opt. Lett. 7, 488 (1998).
[CrossRef]

Proc. SPIE

J. Bruniquel and A. Lopes, Proc. SPIE 2315, 342 (1994).
[CrossRef]

Other

J. W. Goodman, in Laser Speckle and Related PhenomenaJ. C. Dainty, ed., Vol.??9 of Topics in Applied Physics (Springer-Verlag, Berlin, 1975), pp. 9–75.

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (4)

Fig. 1
Fig. 1

(a) Binary support of the 64×64 scene used in the simulation. The target has a size of 737  pixels. (b) Realization of the scene model with parameters μa=3 and μb=1. The initial active contour is superimposed upon this image.

Fig. 2
Fig. 2

Two-dimensional representation of the adopted parameterization of the problem.

Fig. 3
Fig. 3

Comparison of the performance of the monochannel and multichannel methods versus ϕ and δ (see text). Performance is assessed by determination of the average number of misclassified pixels computed for 1000 scene realizations. The error-bar length corresponds to one standard deviation. (a) δ=0.5, (b) δ=0.1, (c) δ=0.0.

Fig. 4
Fig. 4

Three-date intensity one-look SAR image of an agricultural region. (a) 4/30/1993, (b) 6/4/1993, (c) 7/9/1993. Segmentation results are presented for the sum of the three dates. (d) Initialization of the contour. (e) Result obtained with the monochannel method. (f) Result obtained with the multichannel method.

Equations (5)

Equations on this page are rendered with MathJax. Learn more.

we=argmaxw P[w,μa,μb|S],
we=argminw Nawlogl=1Pmalw+Nbwlogl=1Pmblw,
malw=1Nawi|ωi=1sil,mblw=1Nbwi|wi=0sil.
we=argminw Nawl=1Plogmalw+Nbwl=1Plogmblw.
M=μ1+μ22,δ=μ1-μ22.

Metrics